Wikipedia talk:Arbitration/Requests/Case/Rich Farmbrough/Evidence

Irony abounds
HersfoldArbClearkBotClerkBot is counting zero words as one word. Rich Farmbrough, 01:38, 5 April 2012 (UTC).


 * That's a bug I was not aware of, as generally users don't leave blank evidence sections up. The bot will likely be due for an overhaul when the evidence length motions pass, I'll investigate the matter then. It's a known error that the word count can be off by a small number based on minor formatting tweaks, such as extra spacing or excessive newlines, so this is probably an extension of that. It's for this reason the bot's default limits include a 10% tolerance before a warning is issued (that is, you need 550+ words by the bot's count before it yells at you). If you notice any other issues with the bot, I'd appreciate being notified directly. Hers fold  non-admin (t/a/c) 17:27, 5 April 2012 (UTC)


 * Hm. Then again, perhaps this is a bug, it marked my section as having 1 word even though I used one less line break than in your section. I'll have to take a look at this soon. Hers fold  non-admin (t/a/c) 18:04, 5 April 2012 (UTC)


 * I'm not allowed to edit your talk page. Rich Farmbrough, 23:37, 6 April 2012 (UTC).


 * There's always Special:Emailuser/Hersfold. Hers fold  (t/a/c) 02:28, 8 April 2012 (UTC)

Clerking request
Whenaxis' comments do not correspond to the prescribed format. Please remove them. Rich Farmbrough, 01:45, 5 April 2012 (UTC).


 * I see nothing wrong with the format of his evidence. -- Guerillero &#124;  My Talk  13:11, 5 April 2012 (UTC)
 * He makes no assertions in the headers.  Fram does the same.  The pages are set like that for a reason, crystal clarity. and should be something like

 -  -

Example

Rich edits with caps lock on
As seen in these edits where he replaces "obe" with "OBE". 


 * not

Evidence
As seen in these edits where he replaces "obe" with "OBE". 


 * It is vitally important to correct conduct of any case that editors are not allowed to introduce edits that purport to show problems without explicitly stating what the problem is. There are two main reasons:


 * 1) If they are allowed to do that it is impossible to respond effectively to the allegations, for different readers will have different views of what is being alleged.  In the first example above, there might be a simple defence - for example "Not true - see where I replaced "awaded oboe" with "Awarded an OBE."  In the second example a reader might say that such a reply does not explain the useless capitalisation - addressing a completely different issue, that may or may not have been meant.
 * 2) It permits a scatter-gun approach, the throwing in of as much dirt as possible in the hope that some will stick.  We have seen this tactic used repeatedly, to good effect, on Wikipedia, and it should not be allowed to gain a foothold in Arbitration cases.
 * All the best. Rich Farmbrough, 15:04, 5 April 2012 (UTC).


 * To summarise, there is no clarity of purpose, no attempt to show a pattern of behaviour. If Fram merely wants to say "Rich sometimes makes bad edits" then the evidence might be considered as belonging together to attempt to support that statement.  If, on the other hand, he wants to say something more concrete, then the examples that support that particular statement should be presented together. Muddying the waters by spewing a miscellaneous mix of complaint and diff into a couple of vaguely labelled sub-sections is against the very ethos of the procedure which is to bring the harsh light of day onto  what has occurred. Rich Farmbrough, 15:10, 5 April 2012 (UTC).

Rich, I think the notion is "facts speak for themselves". In cases where interpretation is needed, there is the talk page (where we are now), and the "analysis of evidence" workshop section. If there's confusion about what Whenaxis is getting at (and I can see that there might be), it's one thing to ask for explanation here on the talk page, but asking for a clerk to preemptively remove the submission is belligerent and silly. 67.117.145.81 (talk) 16:49, 5 April 2012 (UTC)
 * Just to be on the safe side, I changed my evidence... since Rich requested it. Even though, an Arbitration clerk said it's okay. I find the request a little bizarre though. Regards, Whenaxis (contribs) DR goes to Wikimania! 18:54, 5 April 2012 (UTC)
 * I haven't created an evidence section yet (if I'm going to) as this doesn't seem major enough to merit a rebuttal, but it looks like only the first bullet point in your list is any assertion of a BAG flaw. Could you elaborate? &mdash; madman 19:23, 5 April 2012 (UTC)
 * Thanks for your message, Madman. My original title was a bit ambiguous, I've narrowed it down further. Regards, Whenaxis (contribs) DR goes to Wikimania! 20:23, 5 April 2012 (UTC)
 * I think you mean BRFA flaw, rather than BAG flaw. Headbomb {talk / contribs / physics / books} 15:01, 13 April 2012 (UTC)

Is BRFA not run by BAG? Whenaxis (contribs) DR goes to Wikimania! 15:05, 13 April 2012 (UTC)
 * Yes, but BAG = People, BRFA = Process. Headbomb {talk / contribs / physics / books} 15:23, 13 April 2012 (UTC)
 * I think the people flawed and not the process. Would you not agree? Whenaxis (contribs) DR goes to Wikimania! 18:09, 13 April 2012 (UTC)
 * Then who in the BAG 'flawed'? The first BAG-related thing linked was the BRFA HPB 45, where the bot was approved 'after concerns were brought up that the bot was indefinitely blocked' (in your words). This is nothing special, as malfunctioning bots are indef-blocked all the time (and it was indeed denied while blocked). The bot got unblocked, and the BRFA proceeded with the assumption that the problem were fixed, and task 45 got approved following a successful trial (I'll also note that Task 45 is unrelated to the brouhaha, so denying it would have had zero effect on this situation). The civility problems piled on since the start of the year, the WP:COSMETICBOT issue returned, and task 46 was declined, with task 47-50 being on hold. So I fail to see where the flaw is as far as BAG is concerned, especially since several BAG members tried to reason with RF, and Hersfold (a BAG member) initiated the ARBCOM proceedings. Headbomb {talk / contribs / physics / books} 18:40, 13 April 2012 (UTC)
 * Would the bot being blocked not be a red flag? And the bot was being operated before it was approved not a red flag? I'll consider your points.. and perhaps repeal my evidence regarding that. All the best, Whenaxis (contribs) DR goes to Wikimania! 20:21, 13 April 2012 (UTC)
 * Ok. I've removed that part of my evidence and heading. Thanks for the note. It's about Rich who we're discussing and not BAG who we are talking about so the evidence is useless anyway. Regards, Whenaxis (contribs) DR goes to Wikimania! 00:57, 14 April 2012 (UTC)

desysopping
Other than this bot nonsense, IMHO Rich is a good admin (i.e. above the norm) and it would be a shame to lose his efforts in that area. And as we saw with Betacommand, desysopping by itself doesn't seem likely to stop his troublesome bot operations, other than the few edits involving protected pages, not that big a deal on the scale of things. It's a bit late to ask this, but has anything like mediation been tried? I'd be willing to participate if it could do some good (I commented at some length in Betacommand 3 and some of the RF-related ANI's, and I understand bots pretty well). On a larger scale (across many users), I think we really are at the point where we should address automation abuse by technical means, e.g. by adding a way in MediaWiki to shut off editing API access and/or limit edit rates on a per-user basis, in cases of repeated abuse. The 25 page/day limit suggested in BC3 might be a place to start.

Rich's request concerning Whenaxis does trouble me and seems unlike him (unnecessarily bureaucratic). Maybe he's having a bad day.

67.117.145.81 (talk) 08:27, 5 April 2012 (UTC)
 * Seems unnecessarily bureaucratic to me too; I assume Whenaxis's contributions were made in good faith and can simply be refactored rather than removed entirely. But regarding the technical means you mention, I fail to see how they would be useful at all. A 25 page/day creation limit imposed across the board would be unnecessarily prohibitive to harmless and useful tasks approved by the BAG. A 25 page/day creation limit imposed on individual editors should be imposed by the community, via the BAG, editing restrictions, &c., and at this point ArbCom since all previous efforts have failed. &mdash; madman 13:11, 5 April 2012 (UTC)
 * Obviously an account doing a BAG-approved task to edit more than 25 pages/day would normally not be under a 25 page/day limit. The limit could be turned on and off (or adjusted) on a per-account basis.  It would also be fine to put a rate limit on an editor's main account while still letting them run approved tasks from flagged bot accounts under different limits or no limit.  67.117.145.81 (talk) 16:31, 5 April 2012 (UTC)
 * I doubt any such change would make it into MediaWiki; it's too specialized. The bot user group and the apihighlimits user right already dictate different limits for flagged bot accounts in general (technical means). The bot policy already states that mass page creation tasks (more than 25-50 pages) must be approved by the BAG. If someone engages in unapproved mass page creation, they should be blocked (community means). So either I'm completely misunderstanding your proposal, or the suggested policies and the means to enforce them are already in place. If you're suggesting that ArbCom specifically prohibit Rich Farmbrough from engaging in even approved mass page creation, that's something to discuss in the workshop. &mdash; madman 16:47, 5 April 2012 (UTC)
 * The purpose of a software-enforced limit would be to stop people from running unapproved bots, since community restrictions have repeatedly failed to do that, and bot tasks almost never get reverted (it takes another bot to do that) even if they shouldn't have been done in the first place. There are also incidents where a bot is approved to edit N pages a day and then the operator runs it at much higher speed than was approved (Rich did something like that recently, with his ISBN editing bot).  In cases where a bot is approved at a certain speed, that limit could be set on the bot account.  Obviously approved bots should run under accounts whose limits don't interfere with the approved operation. 67.117.145.81 (talk) 17:02, 5 April 2012 (UTC)
 * I now understand your proposal a bit better. I still think it'd be too specialized (too much complexity and too little use) for inclusion in the MediaWiki core; maybe it'd be suitable for an extension. I also think it would impose way too high a cost of doing business on trustworthy bot operators for too little benefit. Tasks that incite this much community response and tasks that are unapproved are much less common than you seem to think, they tend to be caught quickly, and there are tools for administrators to mass-revert edits. There's a reason community trust is stipulated when bot operation is approved, and I think absent just cause, technical enforcement of hard limits on bots would be contrary to the assumption of good faith of their operators and too great an overreaction to too few incidents. &mdash; madman 17:43, 5 April 2012 (UTC)
 * I know that being the subject of being an arbitration case is very stressful especially as it can lead to desysopping, indefinitely blocks and so on. So, let's deal with Rich in a reasonable way and have sympathy for him. Regards, Whenaxis (contribs) DR goes to Wikimania! 18:56, 5 April 2012 (UTC)
 * It could be considered a bad day that I had just spent hours trying to put together a statement that did meet with the requirements. Being sloppy allows users to just post a pile of diffs with somewhat vaguely damning statements.  This may be fine for  a noticeboard where any old rubbish can be, and often is, posted, but "ARBCOM serious business".  The sloppy attacks over the past few years have shown that, even when they are completely debunked in every respect, they still leave a miasma of bad reputation, and heap of citable sections.  When a sub-page of AN/I was created I was promised that this was not a "badge of shame", and yet it is trotted out as such.  The same with block logs.  I am not process centric, rather the opposite, but there are processes here that are put in place to prevent abuse of the system.  ARBcom is, regardless of intention, pretty much an adversarial process, and pretty much quasi-judicial.  If someone wants to come and post a pile of bot block logs on my talk page, and enquire politely what they are all about, we can have a discussion.  This, however is an evidence phase.  People have made some nasty statements about me, which is fine.  If however they want those statements to be taken seriously (by me at least, and I hope other readers of the case, including Arbitrators) they need to:


 * Have the guts to state the accusation as an assertion.
 * Attempt to stand it up with evidence.
 * So far I see no evidence of this. If anyone does actually decide they want to do this, then I will, perforce be required to spend a significant amount of time making a detailed analysis of what they say.  I am not going to waste more hours upon hours of my time responding to malformed  evidence, which could be removed at any time.
 * I really think that I have a reasonable expectation that process is followed. Rich Farmbrough, 21:06, 5 April 2012 (UTC).


 * Rich, I'm not an admin, thus, I cannot go through your block logs (and your bot's block logs) and find the diff where that occurred. In addition, it's a real pain to go through all of your edits to find evidence to satisfy your criteria you placed above. For your evidence, you provided statements in the same format that I did and used three diffs to support your assertions that is convoluted with opinion. If you propose that ArbCom is "serious business", why did you place an image of 'Antediluvian footwear' that is unrelated to your evidence? Did you think that was funny to mock the claims made by Fram, Elen of the Roads and Hersfold that 'Suspected sockpuppet categories' are harmful to the community? Whenaxis (contribs) DR goes to Wikimania! 21:22, 5 April 2012 (UTC)
 * The bots' block logs are visible in the usual places. 67.119.2.93 (talk) 07:19, 6 April 2012 (UTC)
 * I certainly thought it was funny... 192.251.134.5 (talk) 13:15, 6 April 2012 (UTC)
 * OK lets take those points:
 * Yes I have a dry sense of humour, and it may be considered literally gallows humour in this case. It was Elen of the Roads who remarked that "these are socks dating from before Noah" and while the illustrated pair do not quite meet that criteria, they are the oldest known socks, which is quite good, quite interesting, and quite educational.
 * "Serious business" is an Internet meme. Nonetheless there is much difference between levity and carelessness.
 * You say it's a real pain to go through all of your edits to find evidence to satisfy your criteria you placed above. Well, perhaps then you should not be offering them. As I understood the purpose of ARBCOM is to deal with serious problems, not things that require intense research to even find weak evidence. (And by weak, here, I mean wrong.)  It's not even clear why you are offering evidence, if you have no idea about my past editing history.  Perhaps an amicus brief of some sort?
 * You say we placed statements in the same format. That is true in the sense that it has a header followed by bullet points. But your previous version had no assertion, the current one has four (or three and a random clause) in one header.  That's fine I can work with that, so let's leave it that way.
 * Rich Farmbrough, 23:20, 6 April 2012 (UTC).


 * I came across your case and found it quite interesting. I researched further about you and I thank you for your main faithful contributions to Wikipedia. However, for the well-being of our community, I felt that it was my duty to at least provide some evidence, especially when dealing with Wikipedia's finest, administrators. Please don't take my evidence too harshly and I wish the best of you for the remainder of the case. Regards, Whenaxis (contribs) DR goes to Wikimania! 18:31, 10 April 2012 (UTC)
 * Well then, your contribution is, perhaps, useful. I do rather subscribe to the "adminship is no big deal" school of thought, and think that the quality of Admins is as variable as editors at large, regardless of the increased "toughness" of RfA.  We have lost us not just admins but one or two ARBs to everything from copyvio to paid editing to fraud.  (You might think that with this background people would not get so excited over deleting a few surplus spaces, but that's not human nature - well not on Wikipedia.) Rich Farmbrough, 00:07, 12 April 2012 (UTC).


 * Could I sell you on "bot operation is no big deal" in that case? You're much more valuable to WP as an admin than as a bot op, in my opinion.  That's not especially trying to disparage your bot skills; it's just that there's no shortage of bots on wikipedia at the moment, but rather the opposite. 67.117.147.20 (talk) 07:55, 13 April 2012 (UTC)
 * There may not be a shortage of bots, but one should never spit on any (properly functioning) bot that cleans up articles and assists with backlogs. We could have triple the amount of such bots and we could still use more of them. Headbomb {talk / contribs / physics / books} 14:57, 13 April 2012 (UTC)

Clerking/arbcom request
"Evidence should include a link to the actual page diff in question, or to a short page section; links to the page itself are insufficient. Never link to a page history, an editor's contributions, or a log for all actions of an editor (as those change over time), although a link to a log for a specific article or a specific block log can be useful. Please make sure any page section links are permanent; see simple diff and link guide."

I see both other users linking to pages, and other documents. Something needs to be done about this, since it would create an uneven playing field allowing them to break rules while I attempt to abide by them. (Once again the situation is dripping with irony, considering the accusations against me.)

Rich Farmbrough, 00:23, 7 April 2012 (UTC).


 * The front matter to the evidence page is a guideline prepared for the benefit of the editors submitting evidence; it is not binding. The Committee's official procedures as regards the submission of evidence are to be found at Arbitration Committee/Procedures; they do not constrain the format or target of links, but only their number.
 * More generally, I would suggest focusing on the substance (or lack thereof) of the evidence being presented rather than its format; the latter is unlikely to be the best use of your time. Kirill [talk] [prof] 01:46, 7 April 2012 (UTC)
 * It would be a better use of my time if I knew whether I was allowed to refer to "pages" rather than "diffs" or "small sections", if I knew whether I could refer to a separate set of arguments or comments, just as my detractors are saying "Hey look, here's a whole archive of ****!" I am not fussy what the rules are, but if there are to be rules they should be applied, or if they are not going to be applied, that should be made clear form the start.  The rules should be made clear, and they should be stuck to.  To make rules then allow one party to breach them is a travesty. I am trying to take this case seriously, accused as I am of using WP:IAR, but if it's just a question of "throw as much mud as you can and see what sticks", then it is no different from what Elen of the Roads called "another go round" at AN/I.   Rich Farmbrough, 02:52, 7 April 2012 (UTC).


 * Rich, ignoring rules is a good thing. You're here for repeatedly ignoring other users, not ignoring rules.  My advice about writing your evidence is just remember that it will be read by humans and not bots.  And we all appreciate your intelligence and dedication to this place, even if we think you're going about certain things the wrong way.  So per the policy WP:BURO, don't worry too much about formal procedure in your evidence.  Just make your case straightforwardly, writing for humans who respect you even though they require documented facts and convincing arguments in order to be persuaded.  As you can see from the other presentations, different people used different mixtures of diffs and argument, and the arbs haven't so far complained about it.  67.117.131.84 (talk) 06:44, 10 April 2012 (UTC)

Clarification required
Are rebuttals counted as part of the word count of the evidence? Rich Farmbrough, 00:09, 8 April 2012 (UTC).


 * Yes. Hers fold  (t/a/c) 01:07, 8 April 2012 (UTC)
 * Let me be clearer... Are they supposed to be? This seems prima facie ridiculous. The nature of Arbcom is that one person is often having to defend against  with a bunch of others. Effectively this makes rebuttal well-nigh impossible, which means that the instructions should not even mention it. I have to say I am very disappointed with the process so far, seeing it for the first time as a party, although the actual content is not worrying. Rich Farmbrough, 02:12, 8 April 2012 (UTC).


 * That's partly why we're reviewing and considering revising the limits now. However, cases in the past have included all content on the evidence page as evidence, and thus contributing to the word/diff counts. Again, if you feel someone's evidence is inaccurate/misleading/whatever, the best way to go is likely to provide your own evidence contradicting those claims without directly addressing the other's evidence so as to save space, then compare the two in the Analysis section on the Workshop page. Hers fold  (t/a/c) 02:30, 8 April 2012 (UTC)
 * If you feel another's evidence needs further analysis, however, there is a section specifically for that purpose on the Workshop page. <em style="font-family:Bradley Hand ITC;color:blue">Hers <em style="font-family:Bradley Hand ITC;color:gold">fold  (t/a/c) 01:08, 8 April 2012 (UTC)

fait accompli
I wanted to enter the following into evidence under the heading "fait accompli": However the page is protected. 64.160.39.210 (talk) 17:00, 22 April 2012 (UTC)
 * Rich writes, "It's often easier to do something with AWB than have a BRFA approved. The break-even seems to vary between a few hundred edits and, currently, about 20,000."

I was also thinking of writing up something about Rich's ISBN hyphenation bot. IIRC it was approved based on a representation that it was going to make something like 50 edits a day. It was then apparently inactive for years, then started up again, making 100's or 1000's of edits per day, based on some ANI discussion. I probably won't be able to research the specifics today though, due to other RL stuff that I'm doing. 64.160.39.210 (talk) 17:45, 22 April 2012 (UTC)

Heh, page is actually only semi-protected. 64.160.39.210 (talk) 15:56, 23 April 2012 (UTC)